• DocumentCode
    3408381
  • Title

    How to identify binary relations for domain models

  • Author

    Kaindl, Hermann

  • Author_Institution
    Siemens AG, Vienna, Austria
  • fYear
    1996
  • fDate
    25-29 Mar 1996
  • Firstpage
    28
  • Lastpage
    36
  • Abstract
    Many approaches to requirements engineering include building a model of the domain. Those using entity relationship modeling or deriving from it employ the concept of relations between entities, but identifying the relations is still more of an art than science or engineering. We deal with this problem primarily in the context of object oriented analysis (OOA), where relations between object classes are to be identified. Our new approach uses natural language definitions of object classes and looks for names of other object classes in these definitions, since such a reference indicates a relation. Based on this idea, we identify most binary relations for domain models in a new way. We also provide tool support for this method, which shows that a high degree of automation is possible. Both a case study using the well known ATM (automated teller machine) example and real world experience with our approach suggest its usefulness
  • Keywords
    abstract data types; automatic teller machines; bank data processing; formal specification; natural languages; object-oriented methods; ATM; OOA; automated teller machine; binary relations; domain models; entity relationship modeling; natural language definitions; object classes; object oriented analysis; real world experience; requirements engineering; Art; Automation; Buildings; Context modeling; Data engineering; Databases; Design engineering; Electronic mail; Natural languages; Object oriented modeling;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Software Engineering, 1996., Proceedings of the 18th International Conference on
  • Conference_Location
    Berlin
  • ISSN
    0270-5257
  • Print_ISBN
    0-8186-7247-1
  • Type

    conf

  • DOI
    10.1109/ICSE.1996.493399
  • Filename
    493399